[Mistral75]

Performances as stated and handwritten by Mr Miyazaki (same source as above):

- slight fogging at f/1.1 that is reduced by 70% at f/1.25
- fogging disappears and contrast improves at f/1.4
- between f/1.7 et f/2.8 contrast and resolution keep on improving
- sweet spot is f/4

- vignetting / falloff : 33% at f/1.1 in the image's corners and 7% at f/1.7

- distorsion : +/- 0,3%, at par with a macro lens.

[Clicker]

Thanks Mistral for the proper translation, when i visit japan later this year...i'm gonna be broke coming back

Holy crap i might just get it today!

2012/02/24

09:04


International item released from Customs for processing by Canada Post

[falconeye]

- sweet spot is f/4

This is what I was able to get from the ebay auction images already as posted above.

As I said, sweet spot at f/4 is great for a normal lens (like FourThirds or 35mm Full Frame). But for a P&S sensor as the one in the Q, the sweet spot would have to be at f/2.8 or below for the lens to be a performer. OTOH, other lenses made for larger sensors in this focal range (DA21, Zeiss) are no different.

However, the Nokton 25/0.95 for µFT seems to be better (same performance at f/2 and f/4) as is the Lumix G 20mm f/1.7 ASPH. Latter lenses are likely to outperform this Sonnetar on a Q and don't cost more.

[Mistral75]

Better than the eBay auction images, there is that (same page but better copy).

[deanm3]

Thanks Mistral for the proper translation, when i visit japan later this year...i'm gonna be broke coming back

Holy crap i might just get it today!

2012/02/24

09:04


International item released from Customs for processing by Canada Post

You definitely need to post some impressions. Looks like my Sonnetar is going to get here before my Q by one day, which will be excruciating for me.

[Clicker]

Better than the eBay auction images, there is that (same page but better copy).

good at 1.1-1.4

[Class A]

It provides decent IQ with high ISO and yet you can use f1.1 lens and have the DOF of f5.6. That's nothing short of amazing.

That's not amazing (and no advantage) at all. The impressive "f/1.1" designation purely can be made because the sensor is tiny and the lens hence only has to gather small amounts of lights in order to achieve the same exposure as a lens for an FF sensor.

That statement considers only the total amount of light entering the lens, not the light intensity at the sensor needed for a given exposure which is how the concept of f-stop is used by photographers.

The original question was whether the lens retains the "light gathering" abilities of an "f/1.1" lens on FF (while producing the DOF of a "f/5.6" lens only).

Whether my or your answer is correct depends on how you define "light gathering". You are correct in stating that the lens gathers sufficient light to achieve the same exposure (light density, e.g., cd/mm^2).

I am correct in stating that the total light gathered by the lens is only a fraction of that light that would be gathered by an f/1.1 on an FF sensor. The total light is "sensor-size * light density", hence the lens for the larger sensor needs to gather much more light.

I believe it is wrong to equate "light gathering ability" with "exposure" because what matters for image quality is the total light gathered. That's why the equivalent f-stop for an FF camera is f/5.6, otherwise it would be allowed to collect a lot more light, resulting in much less DOF and more total light. Obviously, f/5.6 results in less exposure which has to be compensated by choosing a higher ISO setting on the FF sensor (-> Falk Lumo: Camera equivalence).

N.B., your thought experiment where parts of the sensor gets masked neglects the fact that the image from the small unmasked part needs to be scaled up more than the original image. If you print your Q images at size 1 x 0.7 (instead of 6 x 4) and look at that tiny print from the same distance as you would look at a 6 x 4 print then you could argue that the resulting quality / light gathering is comparable to the result of an f/1.1 used on FF.

---

Last edited by Class A; 02-24-2012 at 04:21 PM.

[elho_cid]

That's not amazing (and no advantage) at all. The impressive "f/1.1" designation purely can be made because the sensor is tiny and the lens hence only has to gather small amounts of lights in order to achieve the same exposure as a lens for an FF sensor.

But smaller sensors are more efficient. Let's say you have a picture shot at f1.4 with the Q on ISO 1600. It will look acceptable. For the same DOF on a full frame camera you will need ~5 timer higher f-number and ~5 times higher ISO. Mainstream FF camera like 5dmk2 will be noisier. New D1X maybe not, but that's too expensive for most amateurs.

[Class A]

But smaller sensors are more efficient.

This statement contradicts DxOMark results.
The Q has a low-light ISO of 189 while the 5D MkII has ISO 1815. The scores are comparable as they are normalised to a standard 8MP sensor. This advantage would carry through to higher ISO numbers, but due to non-linear behaviour, you have a point that extreme high ISO values on a larger sensor camera may not be noise equivalent to lower ISO values on a smaller sensor.

Let's say you have a picture shot at f1.4 with the Q on ISO 1600. It will look acceptable. For the same DOF on a full frame camera you will need ~5 timer higher f-number and ~5 times higher ISO.

You will need ~25x the ISO setting (not ~5x). Looking at the respective DxOMark comparison, the Canon 5D seems to achieve the Q-ISO 1600 performance at ISO 12800 which is lower than the required ISO 40000. So you have a bit of a point there, but it only applies at very high ISO numbers where the larger sensor of the 5D MkII (for some reason) cannot maintain linearity anymore.

Note that the 5D MkII sensor is outdated while the Q sensor is a very modern design. So the Q sensor is more efficient but not because it is smaller but because it is a more modern design. A comparison with the Nikon D800 would look differently.

[wanderography]

Well, my lens will be here tomorrow most likely... If I can grab the package before noon I'll head out to one of the big cities and do some shooting with it and post the pics...

[Clicker]

Well, my lens will be here tomorrow most likely... If I can grab the package before noon I'll head out to one of the big cities and do some shooting with it and post the pics...

I've been denied a delivery today arrrrgggh looks like it's Monday our Postal system is government run they don't do weekends if they dont have to.

[wanderography]

I've been denied a delivery today arrrrgggh looks like it's Monday our Postal system is government run they don't do weekends if they dont have to.

Ours is government run too, thats why they work weekends lol

[Clicker]

If you get your lens isn't the farmer's Market in Hayward a good shooting opportunity? I dunno but that's up to you.

[wanderography]

If you get your lens isn't the farmer's Market in Hayward a good shooting opportunity? I dunno but that's up to you.

I'm actually staying right across the street from it at my dads house lol... Na, not big around shooting in hayward, too ghetto and too many crack heads and too many people know me...

[elho_cid]

You will need ~25x the ISO setting (not ~5x). Looking at the respective DxOMark comparison, the Canon 5D seems to achieve the Q-ISO 1600 performance at ISO 12800 which is lower than the required ISO 40000.

You are right, I was wrong in my math. I do not know what is the optimum sensor size for light efficiency. I believe it has some link to a CMOS gate size or technology used to build the charge pit for a pixel. But I don't think larger pixels are always an advantage. At some point, mainkg them larger won't help. And as the 90nm technology is now common, the size where increasing size won't help noise efficiency is getting smaller too.
I believe in future of small sensors.
But of course optics is another issue and using smaller than APS-C for large prints is a problem.